Infrastructure for collaboration: Leeds MRC Medical Bioinformatics Centre

For some years medical research has been advanced by new technologies that generate very large amounts of biological data in digital format, for example, new methods that generate very detailed and often three-dimensional images of arthritic joints or new ways of sequencing DNA that bring information on inherited disease. Until recently many of these methods have been limited to research studies, clinical trials and small numbers of patients but recent advances mean that we can now contemplate a future where every individual may know their genome sequence, and where most people will have medical investigations that generate this type of data in routine practice.

In parallel to these developments, the UK NHS is increasingly moving away from using paper to store vital information about patient care and adopting modern computer systems which store patients' health records in a secure electronic format. These patient records are valuable data from which we can learn a great deal about the efficacy of health care and the long term effects of treatments. However, it is clear that the value of these health care records to research will be increased if they can be effectively linked to the large scale data described above. It will be possible for instance to make detailed investigations of how treatments may be better or worse depending on the genetics of the person receiving them, and help move towards a world of personalised medicine.

Bringing large amounts of varied and complex data from different sources is not without challenge. The data need to be accessible and usable by those with a broad range of backgrounds, including researchers familiar with using complex research data and clinicians familiar with accessing patient records. There are also issues around ethics and privacy, plus the development and sustainability of necessary powerful computing infrastructure to support the linking and secure sharing of data between NHS and research institutions.

We aim to overcome these challenges through the creation of the Leeds MRC Medical Bioinformatics Centre which will develop and sustain the infrastructure, experience and facilities needed to undertake ground-breaking research at the point where clinical, health and research data meet. This will be achieved by bringing together a strong team who between them have extensive experience spanning all aspects of the relevant areas including direct patient care, high performance computing, statistics, biomedical and health research. The Centre will build on established and successful links between the University of Leeds, the Leeds Teaching Hospital Trust (LTHT), an external company specialising in medical record systems (The Phoenix Partnership, (TPP)) and the Leeds-based NHS Health & Social Care Information Centre (HSCIC).

The Centre will bring together state of the art technology for visualising data and provide a hub for the training of future researchers in this rapidly developing and complex area. The computing infrastructure will be used to establish a Shared Virtual Datacentre (SVD) which will provide researchers with a highly secure and powerful computing platform.

The Centre will be set up over the first two years of this project, and will support an increasing number of research projects aiming directly at health care improvements. Initial projects will focus on inherited diseases (the genetic causes of neuro-developmental delay), kidney disease (detecting kidney injury before complications ensue), skin cancer (understanding what leads it to spread) and colorectal cancer (understanding the long term effects of surgery combined with radiotherapy). These projects will demonstrate the ways that our Centre will develop new insights into the factors that drive disease, and provide indicators that inform treatment leading to better, more personalised and more effective medicine and health care.

The capital expenditure for the Leeds MRC Medical Bioinformatics Centre will complete an integrated Research Campus linking the University campus to the local Leeds Teaching Hospitals Trust (LTHT), The Phoenix Partnership Ltd (TPP) and HSCIC. This requires upgrades of some existing network links to 100 Gb/s, and extra data storage capability on the University Campus within the Yorkshire and Hull Metropolitan Area Network shared data centre. This is linked to matched funding for infrastructure within LTHT and TPP. Computational processing will be carried out on the University high-performance computing facility, free of direct cost to this application. Capital funds will also be used to refurbish part of a building (Worsley) to house the Centre and associated researchers. Non-capital costs will be met partly from University matched funding and are associated mainly with data scientists and high calibre research fellows. These staff will create and stress test Virtual Research Environment software specialised to the linkage of high-volume molecular (e.g. sequence) data and phenotypic data (e.g. pathology images) to appropriate anonymized health and clinical records. Stress testing will use exemplar projects in disease areas (inherited disease, melanoma, rectal cancer and renal injury) where we already have consented cohorts with significant genomic, proteomic and phenotypic data sets. A key aim of the Centre will be scientific and financial sustainability through the continued creation and funding of new projects based on the infrastructure and capability in place. Other important aspects of the centre will be work on ethics and appropriate electronic consent processes, patient and clinician communication, and engagement and education and training. We will set up and deliver a new MSc programme, a PhD training platform and engage with undergraduate teaching.

We have developed a detailed plan for the achievement of social and economic impact of the MRC Centre which is provided in the 'Pathways to Impact' document. To summarise, we envision a project pathway model (Fig. 2 in the Case for Support) whereby research in medical bioinformatics is driven by clinical questions, and leads to excellent science that goes hand-in-hand with ultimate economic, societal and/or clinical impact. We have put in place a range of activities to support this model, and achieve the desired impacts. These include regular seminars, conferences and workshops addressing clinicians, industry and policy makers where the Centre's achievements, its methodology and its projects will be disseminated.

We recognise the key role to be played in our processes by the patients whose records and data we will use to drive research. It is critical to the impact of the Centre that large numbers of patients understand how their anonymised data can lead to research advances of societal and economic benefit, and that they consent to this. We will promote this through regular patient engagement meetings, and making the consent processes and software we will develop during the project as clear and user-friendly as possible.

Our communication plans are also described elsewhere. They include a highly functional, content-rich website, e-newsletters, social media and other communications activities. These will reach a diverse and wide range of stakeholders with messages in appropriate language tailored their interests.

Training and capacity development will be a major part of the Centre's activities, from undergraduates in medicine and other disciplines, Masters students on our subject specific courses, a large cohort of PhD students, research fellows and other professionals engaging in CPD activities (our modules are usually available in this format). This will produce a large number of well-trained professionals who will provide impact from academic research to small biotechnology companies and larger pharmaceutical companies.